inet regression test 
Author Message
 inet regression test

--=-J2AKRatNSCjSeULgVk77
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

Looks like some ORDER BY statements would be useful.

*** ./expected/inet.out Sat Jun 16 22:05:20 2001
--- ./results/inet.out  Wed Jan 15 10:18:40 2003
***************
*** 193,212 ****
  select * from inet_tbl where i<<'192.168.1.0/24'::cidr;
         c        |        i=20=20=20=20=20=20=20=20=20
  ----------------+------------------
   192.168.1.0/24 | 192.168.1.0/25
   192.168.1.0/24 | 192.168.1.255/25
-  192.168.1.0/24 | 192.168.1.226
  (3 rows)
=20=20
  select * from inet_tbl where i<<=3D'192.168.1.0/24'::cidr;
         c        |        i=20=20=20=20=20=20=20=20=20
  ----------------+------------------
-  192.168.1.0/24 | 192.168.1.0/24
   192.168.1.0/24 | 192.168.1.226/24
!  192.168.1.0/24 | 192.168.1.255/24
   192.168.1.0/24 | 192.168.1.0/25
   192.168.1.0/24 | 192.168.1.255/25
-  192.168.1.0/24 | 192.168.1.226
  (6 rows)
=20=20
  set enable_seqscan to on;
--- 193,212 ----
  select * from inet_tbl where i<<'192.168.1.0/24'::cidr;
         c        |        i=20=20=20=20=20=20=20=20=20
  ----------------+------------------
+  192.168.1.0/24 | 192.168.1.226
   192.168.1.0/24 | 192.168.1.0/25
   192.168.1.0/24 | 192.168.1.255/25
  (3 rows)
=20=20
  select * from inet_tbl where i<<=3D'192.168.1.0/24'::cidr;
         c        |        i=20=20=20=20=20=20=20=20=20
  ----------------+------------------
   192.168.1.0/24 | 192.168.1.226/24
!  192.168.1.0/24 | 192.168.1.226
!  192.168.1.0/24 | 192.168.1.0/24
   192.168.1.0/24 | 192.168.1.0/25
+  192.168.1.0/24 | 192.168.1.255/24
   192.168.1.0/24 | 192.168.1.255/25
  (6 rows)
=20=20
  set enable_seqscan to on;

--=20

PGP Key: http://www.***.com/

--=-J2AKRatNSCjSeULgVk77
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)

iD8DBQA+JYbE6DETLow6vwwRApAwAJ9XymX+cJVy6W2/dJrNtCJmy3hihwCfV+5X
/1GeOZEeB1i0j4dtgco6MPk=
=45Ov
-----END PGP SIGNATURE-----

--=-J2AKRatNSCjSeULgVk77--



Mon, 04 Jul 2005 00:21:01 GMT
 inet regression test

--ELM1042648530-2851-1_
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset=US-ASCII

OK, I have applied an patch to ORDER BY on those queries. I don't see
the ordering you have here, but this should fix it for you.  I also
uppercased the keywords while I was in there.

---------------------------------------------------------------------------

-- Start of PGP signed section.

Quote:
> Looks like some ORDER BY statements would be useful.

> *** ./expected/inet.out    Sat Jun 16 22:05:20 2001
> --- ./results/inet.out     Wed Jan 15 10:18:40 2003
> ***************
> *** 193,212 ****
>   select * from inet_tbl where i<<'192.168.1.0/24'::cidr;
>          c        |        i        
>   ----------------+------------------
>    192.168.1.0/24 | 192.168.1.0/25
>    192.168.1.0/24 | 192.168.1.255/25
> -  192.168.1.0/24 | 192.168.1.226
>   (3 rows)

>   select * from inet_tbl where i<<='192.168.1.0/24'::cidr;
>          c        |        i        
>   ----------------+------------------
> -  192.168.1.0/24 | 192.168.1.0/24
>    192.168.1.0/24 | 192.168.1.226/24
> !  192.168.1.0/24 | 192.168.1.255/24
>    192.168.1.0/24 | 192.168.1.0/25
>    192.168.1.0/24 | 192.168.1.255/25
> -  192.168.1.0/24 | 192.168.1.226
>   (6 rows)

>   set enable_seqscan to on;
> --- 193,212 ----
>   select * from inet_tbl where i<<'192.168.1.0/24'::cidr;
>          c        |        i        
>   ----------------+------------------
> +  192.168.1.0/24 | 192.168.1.226
>    192.168.1.0/24 | 192.168.1.0/25
>    192.168.1.0/24 | 192.168.1.255/25
>   (3 rows)

>   select * from inet_tbl where i<<='192.168.1.0/24'::cidr;
>          c        |        i        
>   ----------------+------------------
>    192.168.1.0/24 | 192.168.1.226/24
> !  192.168.1.0/24 | 192.168.1.226
> !  192.168.1.0/24 | 192.168.1.0/24
>    192.168.1.0/24 | 192.168.1.0/25
> +  192.168.1.0/24 | 192.168.1.255/24
>    192.168.1.0/24 | 192.168.1.255/25
>   (6 rows)

>   set enable_seqscan to on;

> --

> PGP Key: http://www.rbt.ca/rbtpub.asc

-- End of PGP section, PGP failed!

--
  Bruce Momjian                        |  http://candle.pha.pa.us

  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

--ELM1042648530-2851-1_
Content-Transfer-Encoding: 7bit
Content-Type: text/plain
Content-Disposition: inline; filename="/bjm/diff"

Index: src/test/regress/expected/inet.out
===================================================================
RCS file: /cvsroot/pgsql-server/src/test/regress/expected/inet.out,v
retrieving revision 1.13
diff -c -c -r1.13 inet.out
*** src/test/regress/expected/inet.out  17 Jun 2001 02:05:20 -0000      1.13
--- src/test/regress/expected/inet.out  15 Jan 2003 16:34:09 -0000
***************
*** 168,174 ****
  (14 rows)

  -- check the conversion to/from text and set_netmask
! select '' AS ten, set_masklen(inet(text(i)), 24) FROM INET_TBL;
   ten |   set_masklen    
  -----+------------------
       | 192.168.1.226/24
--- 168,174 ----
  (14 rows)

  -- check the conversion to/from text and set_netmask
! SELECT '' AS ten, set_masklen(inet(text(i)), 24) FROM INET_TBL;
   ten |   set_masklen    
  -----+------------------
       | 192.168.1.226/24
***************
*** 188,196 ****
  (14 rows)

  -- check that index works correctly
! create index inet_idx1 on inet_tbl(i);
! set enable_seqscan to off;
! select * from inet_tbl where i<<'192.168.1.0/24'::cidr;
         c        |        i        
  ----------------+------------------
   192.168.1.0/24 | 192.168.1.0/25
--- 188,196 ----
  (14 rows)

  -- check that index works correctly
! CREATE INDEX inet_idx1 ON inet_tbl(i);
! SET enable_seqscan TO off;
! SELECT * FROM inet_tbl WHERE i<<'192.168.1.0/24'::cidr ORDER BY c,i;
         c        |        i        
  ----------------+------------------
   192.168.1.0/24 | 192.168.1.0/25
***************
*** 198,204 ****
   192.168.1.0/24 | 192.168.1.226
  (3 rows)

! select * from inet_tbl where i<<='192.168.1.0/24'::cidr;
         c        |        i        
  ----------------+------------------
   192.168.1.0/24 | 192.168.1.0/24
--- 198,204 ----
   192.168.1.0/24 | 192.168.1.226
  (3 rows)

! SELECT * FROM inet_tbl WHERE i<<='192.168.1.0/24'::cidr ORDER BY c,i;
         c        |        i        
  ----------------+------------------
   192.168.1.0/24 | 192.168.1.0/24
***************
*** 209,213 ****
   192.168.1.0/24 | 192.168.1.226
  (6 rows)

! set enable_seqscan to on;
! drop index inet_idx1;
--- 209,213 ----
   192.168.1.0/24 | 192.168.1.226
  (6 rows)

! SET enable_seqscan TO on;
! DROP INDEX inet_idx1;
Index: src/test/regress/sql/inet.sql
===================================================================
RCS file: /cvsroot/pgsql-server/src/test/regress/sql/inet.sql,v
retrieving revision 1.7
diff -c -c -r1.7 inet.sql
*** src/test/regress/sql/inet.sql       17 Jun 2001 02:05:20 -0000      1.7
--- src/test/regress/sql/inet.sql       15 Jan 2003 16:34:10 -0000
***************
*** 52,63 ****
    FROM INET_TBL;

  -- check the conversion to/from text and set_netmask
! select '' AS ten, set_masklen(inet(text(i)), 24) FROM INET_TBL;
  -- check that index works correctly
! create index inet_idx1 on inet_tbl(i);
! set enable_seqscan to off;
! select * from inet_tbl where i<<'192.168.1.0/24'::cidr;
! select * from inet_tbl where i<<='192.168.1.0/24'::cidr;
! set enable_seqscan to on;
! drop index inet_idx1;

--- 52,63 ----
    FROM INET_TBL;

  -- check the conversion to/from text and set_netmask
! SELECT '' AS ten, set_masklen(inet(text(i)), 24) FROM INET_TBL;
  -- check that index works correctly
! CREATE INDEX inet_idx1 ON inet_tbl(i);
! SET enable_seqscan TO off;
! SELECT * FROM inet_tbl WHERE i<<'192.168.1.0/24'::cidr ORDER BY c,i;
! SELECT * FROM inet_tbl WHERE i<<='192.168.1.0/24'::cidr ORDER BY c,i;
! SET enable_seqscan TO on;
! DROP INDEX inet_idx1;

--ELM1042648530-2851-1_
Content-Type: text/plain
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
MIME-Version: 1.0

---------------------------(end of broadcast)---------------------------
TIP 5: Have you checked our extensive FAQ?

http://www.postgresql.org/users-lounge/docs/faq.html

--ELM1042648530-2851-1_--



Mon, 04 Jul 2005 01:05:55 GMT
 inet regression test

Quote:

> Looks like some ORDER BY statements would be useful.

To do what?  Those queries should already be producing indexscan plans.
If you're not getting the expected answers, there is something that
needs to be fixed, not papered over.

Please try

create index inet_idx1 on inet_tbl(i);
set enable_seqscan to off;
explain select * from inet_tbl where i<<'192.168.1.0/24'::cidr;

                        regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 3: if posting/reading through Usenet, please send an appropriate

message can get through to the mailing list cleanly



Mon, 04 Jul 2005 04:54:54 GMT
 inet regression test

If you get it working, I can remove the ORDER BY's I added.

---------------------------------------------------------------------------

Quote:


> > Looks like some ORDER BY statements would be useful.

> To do what?  Those queries should already be producing indexscan plans.
> If you're not getting the expected answers, there is something that
> needs to be fixed, not papered over.

> Please try

> create index inet_idx1 on inet_tbl(i);
> set enable_seqscan to off;
> explain select * from inet_tbl where i<<'192.168.1.0/24'::cidr;

>                    regards, tom lane

--
  Bruce Momjian                        |  http://candle.pha.pa.us

  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

---------------------------(end of broadcast)---------------------------



Mon, 04 Jul 2005 04:56:50 GMT
 inet regression test
--=-xFXNhk3wqi89zuIKGRGf
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

regression=3D#=20
regression=3D#=20
regression=3D# create index inet_idx1 on inet_tbl(i);
CREATE INDEX
regression=3D# set enable_seqscan to off;
SET
regression=3D# explain select * from inet_tbl where
i<<'192.168.1.0/24'::cidr;
                                  QUERY
PLAN=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=
=20=20=20=20=20=20=20=20=20=20=20=20
---------------------------------------------------------------------------=
----
 Index Scan using inet_idx1 on inet_tbl  (cost=3D0.00..4.68 rows=3D7
width=3D64)
   Index Cond: ((i > '192.168.1.0/24'::inet) AND (i <=3D
'192.168.1.255'::inet))
   Filter: (i << '192.168.1.0/24'::inet)
(3 rows)

regression=3D# select * from inet_tbl where i<<'192.168.1.0/24'::cidr;
       c        |        i=20=20=20=20=20=20=20=20=20
----------------+------------------
 192.168.1.0/24 | 192.168.1.0/25
 192.168.1.0/24 | 192.168.1.255/25
 192.168.1.0/24 | 192.168.1.226
(3 rows)

Quote:


> > Looks like some ORDER BY statements would be useful.
>=20
> To do what?  Those queries should already be producing indexscan plans.
> If you're not getting the expected answers, there is something that
> needs to be fixed, not papered over.
>=20
> Please try
>=20
> create index inet_idx1 on inet_tbl(i);
> set enable_seqscan to off;
> explain select * from inet_tbl where i<<'192.168.1.0/24'::cidr;
>=20
>                    regards, tom lane

--=20

PGP Key: http://www.rbt.ca/rbtpub.asc

--=-xFXNhk3wqi89zuIKGRGf
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)

iD8DBQA+JcJM6DETLow6vwwRAu46AJ9rqhr94rvqxkC1Cw6mJNGEsM337QCfaAmg
VCMysxCT1DyaQsdfZA3DI60=
=9TyY
-----END PGP SIGNATURE-----

--=-xFXNhk3wqi89zuIKGRGf--



Mon, 04 Jul 2005 05:11:28 GMT
 inet regression test
--=-7oHr6p8Of9BSf6o3uaD9
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

With Bruce's patch came a few others -- plus a recompile with all
updates.

Quote:
> ... and that's the expected result.  So why'd you get a different result
> while running the regression test?  Curiouser and curiouser...

--=20

PGP Key: http://www.rbt.ca/rbtpub.asc

--=-7oHr6p8Of9BSf6o3uaD9
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)

iD8DBQA+JcQk6DETLow6vwwRAu0yAJwKQrvTzCNEocTcMUnZqMBHiQXnMgCeIHZg
qwzDYV/rZUL1Pt9+xdDJkZA=
=8IaT
-----END PGP SIGNATURE-----

--=-7oHr6p8Of9BSf6o3uaD9--



Mon, 04 Jul 2005 05:21:22 GMT
 inet regression test
--=-z6Ad2rb1kAcvpVi2Yd0t
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

Quote:


> > With Bruce's patch came a few others -- plus a recompile with all
> > updates.
>=20
> Which patch exactly?  The regression test was still passing for me
> as of yesterday's sources plus the large planner commit I just made.
> I'm resyncing to CVS tip at the moment ...

Last update was about 5 minutes after Bruce's header fix went into
place.

--=20

PGP Key: http://www.rbt.ca/rbtpub.asc

--=-z6Ad2rb1kAcvpVi2Yd0t
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)

iD8DBQA+JcUc6DETLow6vwwRAmaHAJ44fMhAzUVec6ecaF4IMjyxYdKUPgCeKywa
KBStiroYGehTkvyf/4gE9S4=
=rDlm
-----END PGP SIGNATURE-----

--=-z6Ad2rb1kAcvpVi2Yd0t--



Mon, 04 Jul 2005 05:21:56 GMT
 inet regression test

Quote:

> With Bruce's patch came a few others -- plus a recompile with all
> updates.

Which patch exactly?  The regression test was still passing for me
as of yesterday's sources plus the large planner commit I just made.
I'm resyncing to CVS tip at the moment ...

                        regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 4: Don't 'kill -9' the postmaster



Mon, 04 Jul 2005 05:22:28 GMT
 inet regression test

Quote:

>  Index Scan using inet_idx1 on inet_tbl  (cost=3D0.00..4.68 rows=3D7
> width=3D64)
>    Index Cond: ((i > '192.168.1.0/24'::inet) AND (i <=3D
> '192.168.1.255'::inet))
>    Filter: (i << '192.168.1.0/24'::inet)
> (3 rows)

That's the expected plan ...

Quote:
> regression=3D# select * from inet_tbl where i<<'192.168.1.0/24'::cidr;
>        c        |        i        =20
> ----------------+------------------
>  192.168.1.0/24 | 192.168.1.0/25
>  192.168.1.0/24 | 192.168.1.255/25
>  192.168.1.0/24 | 192.168.1.226
> (3 rows)

... and that's the expected result.  So why'd you get a different result
while running the regression test?  Curiouser and curiouser...

                        regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?

http://archives.postgresql.org



Mon, 04 Jul 2005 05:29:00 GMT
 inet regression test

Quote:

> Last update was about 5 minutes after Bruce's header fix went into
> place.

Hm.  I just finished verifying that CVS tip builds and passes cleanly
on both HPUX and Linux (Red Hat 8.0).  So either you've got a build
error (did you do a "make clean" after your last update?) or there's
some really weird platform dependency involved.  What's your platform
again?

                        regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 4: Don't 'kill -9' the postmaster



Mon, 04 Jul 2005 06:27:37 GMT
 inet regression test
--=-lQp1elzlyYkl3Wg00JXz
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

Quote:


> > Last update was about 5 minutes after Bruce's header fix went into
> > place.
> Hm.  I just finished verifying that CVS tip builds and passes cleanly
> on both HPUX and Linux (Red Hat 8.0).  So either you've got a build
> error (did you do a "make clean" after your last update?) or there's
> some really weird platform dependency involved.  What's your platform
> again?

I do a make distclean.

FreeBSD 4.7

--=20

PGP Key: http://www.rbt.ca/rbtpub.asc

--=-lQp1elzlyYkl3Wg00JXz
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)

iD8DBQA+Jc+z6DETLow6vwwRAlbHAJ9EpMvBNBcBxgJerBaZ++Z6VOqtZgCfS2Sm
9Xocr5hy/Q6XPlodik+zIgw=
=PFRd
-----END PGP SIGNATURE-----

--=-lQp1elzlyYkl3Wg00JXz--



Mon, 04 Jul 2005 07:59:51 GMT
 inet regression test

Quote:


>> Hm.  I just finished verifying that CVS tip builds and passes cleanly
>> on both HPUX and Linux (Red Hat 8.0).  So either you've got a build
>> error (did you do a "make clean" after your last update?) or there's
>> some really weird platform dependency involved.  What's your platform
>> again?
> I do a make distclean.
> FreeBSD 4.7

I'm still not able to duplicate any problem.  Any other FreeBSD folk see
inet regression failures in CVS tip?

                        regards, tom lane

---------------------------(end of broadcast)---------------------------



Mon, 04 Jul 2005 09:58:21 GMT
 inet regression test

Quote:
> >> Hm.  I just finished verifying that CVS tip builds and passes cleanly
> >> on both HPUX and Linux (Red Hat 8.0).  So either you've got a build
> >> error (did you do a "make clean" after your last update?) or there's
> >> some really weird platform dependency involved.  What's your platform
> >> again?

> > I do a make distclean.

> > FreeBSD 4.7

> I'm still not able to duplicate any problem.  Any other FreeBSD folk see
> inet regression failures in CVS tip?

Unable to reproduce this on 5.0 with PostgreSQL HEAD as of 8:10PST.
-sc

--
Sean Chittenden

---------------------------(end of broadcast)---------------------------
TIP 3: if posting/reading through Usenet, please send an appropriate

message can get through to the mailing list cleanly



Mon, 04 Jul 2005 12:11:16 GMT
 inet regression test
--=-dBTa0lbswsiinG0DB68h
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

Quote:



> >> Hm.  I just finished verifying that CVS tip builds and passes cleanly
> >> on both HPUX and Linux (Red Hat 8.0).  So either you've got a build
> >> error (did you do a "make clean" after your last update?) or there's
> >> some really weird platform dependency involved.  What's your platform
> >> again?
>=20
> > I do a make distclean.
>=20
> > FreeBSD 4.7
>=20
> I'm still not able to duplicate any problem.  Any other FreeBSD folk see
> inet regression failures in CVS tip?

Comes out fine now (~ 12:30 EST5EDT). I'll keep an eye out, but it must
be something else I was doing at the time that threw it off (compiler
glitch?)

Ran the below:

make distclean > /dev/null
cvs update -dP

CFLAGS=3D"-O2 -Wall -Wmissing-prototypes -Wmissing-declarations"
export CFLAGS

PGDIR=3D`pwd | sed -E 's/(.*)\/pgsql([^\/]+)$/db\2/g'`
export PGDIR

./configure --prefix=3D${HOME}/work/postgresql/${PGDIR} \
            --enable-depend \
            --enable-debug \
            --enable-cassert \
            --with-pgport=3D`perl -e 'print int(rand(a) * 999) + 6000;'`

make install > /dev/null
make check

--=20

PGP Key: http://www.rbt.ca/rbtpub.asc

--=-dBTa0lbswsiinG0DB68h
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (FreeBSD)

iD8DBQA+JkZT6DETLow6vwwRAojnAJ0ZCR0hYLbIfGQjml8Ti2HykXVJ9wCdG+84
XhdFjfYCLawnYnTpaCoKzPw=
=4AZ1
-----END PGP SIGNATURE-----

--=-dBTa0lbswsiinG0DB68h--



Mon, 04 Jul 2005 13:42:49 GMT
 
 [ 14 post ] 

 Relevant Pages 

1. Another regression test fails to stand the test of time

2. Regression tests failing on abstime due to test date being

3. TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST TEST

4. inconsistant regression test results...

5. Regression test drivers

6. tsearch - Regression tests fail

7. regression test

8. Regression test questions.

9. Geometry regression tests (was Re: Float output formatting

10. USE_AUSTRALIAN_RULES breaks regression tests in 7.1.2?

11. Is the 7.3.1 geometry regression test supposed to pass perfectly

12. Regression test fails


 
Powered by phpBB® Forum Software